MICHAEL EDWARDS 39 Have NGOs 'Made a Difference?' From Manchester to Birmingham with an Elephant in the Room Michael Edwards Id 1991, David Hulme and 1 found ourselves in a bar at the University of Hull enjoying a post-conference beer.1 The conversation turned to a mutual interest of ours — the role and impact of NGOs in development - and after a few more pints we hit on the idea that eventually became the first 'Manchester Conference* on the theme of 'scaling-up\ later to be summarized in a book titled Making a Difference: NGOs and Development in a Changing World (Edwards and Hulme, 1992). Fifteen years on, the NGO universe has been substantially transformed, with rates of growth in scale and profile that once would have been unthinkable. Yet still the nagging questions remain. Despite the increasing size and sophistication of the development NGO sector, have NGOs really 'made a difference' in the ways the first Manchester Conference intended, or have the reforms that animated the NGO community during the 1990s now run out of steam? In this chapter I try to answer these questions in two ways. First, through a retrospective look at the Manchester conferences — what they taught us, what influence they had, and how NGOs have changed. And second, by picking out a couple of especially important challenges in development terms and assessing whether NGOs 'stood up to be counted', so to speak, and did their best in addressing them. These two approaches suggest somewhat different conclusions, which will bring me to the 'elephant in the room' of my title. It is obvious that making judgements about a universe as diverse as ( development NGOs is replete with dangers of overgeneralization, and difficulties of attribution, measurement, context and timing. I suspect that my conclusions may be particularly relevant for international NGOs and to larger intermediary NGOs based in the South. So, with these caveats in mind, what does the last decade and a half tell us about the role and impact of NGOs in development? The Manchester Conferences: A Short Retrospective As Table 3.1 shows, the theme of the first Manchester Conference in 1992 was 'Scaling-up NGO impact on development: how can NGOs progress from improving local situations on a small scale to influencing the wider systems that create and reinforce poverty?' (Edwards and Hulme, 1992: 7). The conference concluded that there were different strategies suited to different circumstances, specifically: (1) working with government; (2) operational expansion; (3) lobbying and advocacy; {4) and networking and 'self-spreading' local initiatives. All of these strategies have costs and benefits, but the implicit bias of the conference organizers, and most of the participants, lay towards institutional development and advocacy as the most effective and least costly forms of scaling-up, what Alan Fowler later called the 'onion-skin strategy for NGOs — a solid core of concrete practice (either direct project implementation or support to other organizations and their work), surrounded by successive and interrelated layers of research and evaluation, advocacy and campaigning, and public education. To varying extents, this strategy has become standard practice for development NGOs in the intervening years. Buried away at the end of Making a Difference was the following statement: 'The degree to which a strategy or mix of strategies compromises the logic by which legitimacy is claimed provides a useful test of whether organizational self-interest is subordinating mission' (Edwards and Hulme, 1992: 213). For reasons that I will come back to later in my argument, that has turned out to be a prescient conclusion. Fast-forward to the second Manchester Conference in 1994, in a context in which NGOs had begun to 'scale-up' rapidly in an environment in which they were seen as important vehicles to deliver the political and economic objectives of the 'New Policy Agenda' that was being adopted by official donor agencies at the time — deeper democratization through the growth of 'civil society', and more cost-effective delivery of development-related services such as micro-credit and community-driven development. As a result, many NGO budgets were financed increasingly by government aid, raising critical questions about performance, accountability and relations with funding sources. The key question for that conference was as follows: 'Will NGOs be co-opted into the New Policy Agenda as the favored child, or magic bullet for development?' (Edwards and Hulme, 1995: 7). And, if so, what would that do to NGO mission and relationships? Will they, as Table 2.1 The Manchester conferences: a summary Location and date Theme (s) Manchester Scaling-up NGO impact on 1992 development: 'How can NGOs progress from improving local situations on a small scale to influencing the wider systems that create and reinforce poverty?' Key conclusions Published outputs Different strategies suit different circumstances: (1) working with government; (2) operational expansion; (3) lobbying and advocacy; (4) networking and 'self-spreading' local initiatives. All have costs and benefits but implicit bias to institutional development and advocacy to control for dangers (the (onion-skin' strategy): 'The degree to which a strategy or mix of strategies compromises the logic by which legitimacy is claimed provides a useful test of whether organizational self-interest is subordinating Making a Difference: NGOs and Development in a Changing World Scaling-up NGO Impact on Development: Learning from Experience (DIP) Manchester NGO growth raises questions about 1994 performance, accountability and relations with funding sources: • 'Will NGOs be co-opted into the New Policy Agenda as the favored child, or magic bullet for development?' • If so, what does that do to NGO mission and relationships: 'too close to the powerful, too far from the powerless'? Problems are not inevitable - they depend on the quality of relationships between actors and how 'room to manoeuvre' is exploited. Therefore, negotiation between stakeholders is vital, requiring innovation in performance assessment, accountability mechanisms, and relations with funders. 'The developmental impact of NGOs, their capacity to attract support, and their legitimacy as actors in development, will rest much more clearly on their ability to demonstrate that they can perform effectively and are accountable for their actions. It is none to soon for NGOs to put their house in order.' Beyond the Magic Bullet: NGO Performance and Accountability in the Post Cold-War World (x 2) NGOs, States and Donors: Too Close for Comfort? (x 2) Too Close For Comfort: The Impact of Official Aid on NGOs (WD) Policy Arena: New Roles and Challenges for NGOs (JID) Birmingham The changing global context 1999 poses questions about NGO roles, relationships, capacities and accountabilities. 'Adapt or die!' Three key changes: 1. globalization reshapes patterns of poverty, inequality and insecurity; 2. 'complex political emergencies' reshape humanitarian action; 3. the focus of international cooperation is moving from foreign aid to rules, standards and support for the most vulnerable. Hence transnational organizing among equals for systemic change in North-South transfers and interventions. Manchester NGOs and development alternatives: 2005 have we really changed things? NGOs have helped to change the debate on globalization, increase commitment to participation and human rights, and keep the spotlight on the need for reforms in the international system (trade, intervention etc.). But the foreign aid system/paradigm has changed much less than was predicted in 1999. Has this been a disincentive to deeper changes in NGO practice (the 'security blanket' effect)? NGOs in a Global Future: Marrying Local Delivery to Worldwide Leverage (PAD) New Roles and Relevance: Development NGOs and the Challenge of Change NGO Futures: Beyond Aid (TWQ) Global Citizen Action This changing context gives rise to four challenges for NGOs: 6 1. mobilizing a genuinely inclusive civil society at all levels of the world system; 2. holding other organizations accountable for their actions and ensuring they respond to social and environmental needs; 3. ensuring that international regimes are implemented effectively and to the benefit of poor countries; 4. ensuring that gains at the global level are translated into concrete benefits at the grassroots. NGOs must move from 'development as delivery to development as leverage', or marry local development J to worldwide leverage'. This requires more equal relationships with other civic actors, especially in the South, new capacities (e.g. bridging and mediation), and stronger accountability mechanisms. Significant changes in the external environment: NGOs and the Challenge • increasing pace of global change and commonality in of Development causes and effects (no more 'North' and 'South'?); Alternatives • geopolitical rearrangements and their impact on global Have NGOs 'Made a governance (USA, China, India /Brazil/South Africa, Difference'? Middle East); ' From Manchester to ' cultural cleavages on values and ideology (religion); Birmingham with an • the reality of climate change, esp. given urbanization. Elephant in the Room But also stronger conventional international cooperation (increased ODA; continued donor influence, imposed democratization and economic reform, democratic deficits in international institutions, despite recipients' dissatisfaction and growing external criticism). Will the international system, including NGOs, change faced with new global realities? 42 CAN NGOS MAKE A DIFFERENCE? another of the conference books put it (Hulme and Edwards, 1097; 275), become 'too close to the powerful, and too far from the powerless'? At the time, our conclusion was that such problems were not inevitable. Whether they arise depends on the quality of the relationships that develop between actors, and on how each NGO uses its 'room-to-manoeuvre' to control for the costs of growth and donor-dependence. Therefore, negotiation between stakeholders is vital, requiring innovation in performance assessment, accountability mechanisms, and relations with funding agencies. 'The developmental impact of NGOs,' we concluded, 'their capacity to attract support, and their legitimacy as actors in development, will rest much more clearly on their ability to demonstrate that they can perform effectively and are accountable for their actions. It is none too soon for NGOs to put their house in order' (Edwards and Hulme, 1995: 227-8). Since 1994 there have been some important innovations in this respect, \ like the Humanitarian Accountability Project; the rise of self-certification and accreditation schemes, seals of approval and codes of conduct among child sponsorship agencies and other NGOs; the development of formal compacts between government and the non-profit sector in the UK, Canada and elsewhere; the Global Accountability Project in London; ActionAid's ALNAP system; and simple but powerful things like publicizing the financial accounts of an NGO on public bulletin boards that are being encouraged by MANGO and other organizations (Jordan and van Tuijl, 2006). In retrospect, however, NGOs did not heed this call with sufficient attention, and are now suffering from it in a climate in which, unlike ten ! years ago, weaknesses in NGO accountability are being used as cover for I an attack on political grounds against voices that certain interests wish to silence. Examples of such attacks include the NGO Watch project at the American Enterprise Institute, the Rushford Report in Washington DC, and NGO Monitor in Jerusalem. Stronger NGO accountability mechanisms ^ won't do away with politically motivated attacks like these, but they would surely help to expose them for what they are. In 1999, the Third NGO Conference took place in Birmingham, framed by a rapidly changing global context that posed some deeper questions about NGO roles, relationships, capacities and accountabilities. 'Adapt or die' was the subtext of that meeting, whose organizers highlighted three key sets of changes: First, globalization reshapes patterns of poverty, inequality and insecurity, calling for greater global integration of NGO strategies and more 'development work' of different kinds in the North; Second, 'complex political emergencies' reshape patterns of humanitarian action, implying more difficult choices for NGOs about intervention and the need to re-assert their independence from government interests; and, MICHAEL EDWARDS 43 Third, a move from foreign aid as the key driver of international cooperation to a focus on rules, standards and support for those who are most vulnerable to the negative effects of global change implies greater NGO involvement in the processes and institutions of global governance, both formal and informal. (Edwards et at, 1099: 2) The thrust of these changes is clearly visible in the titles of the books that emerged from the Birmingham conference - NGO Futures: Beyond Aid (Fowler, 2000); New Roles and Relevance (Lewis and Wallace, 2000); and Global Citizen Action (Edwards and Gaventa, 2001) - holding out the promise of transnational organizing among equals for systemic change as opposed to a secondary role shaped by the continued asymmetries of the foreign aid world. This changing context, we believed, gave rise to four key challenges resulting from the evolution of a more political role for development NGOs in emerging systems of global governance, debate and decision making: 1. how to mobilize a genuinely inclusive civil society at all levels of the world system, as opposed to a thin layer of elite NGOs operating internationally; 2. how to hold other (more powerful) organizations accountable for their actions and ensure that they respond to social and environmental needs — something that implicitly demanded reforms in NGO accountability; 3. How to ensure that international regimes are implemented effectively and to the benefit of poor people and poor countries (getting to grips with 'democratic deficits' in global institutions and protecting 'policy space' for Southern countries to embark on their own development strategies); and 4. how to ensure that gains at the global level are translated into concrete benefits at the grassroots, translating abstract commitments made in international conferences into actions that actually enforce rules and regulations on the ground (Edwards et al., 1999: 10). ! NGOs, we concluded, must move from 'development as delivery to development as leverage*, and this would require the development of more equal relationships with other civic actors, especially in the South, new capacities (like bridging and mediation), and stronger downward or horizontal accountability mechanisms. Since 1999 there have certainly been some examples of innovations like these, like the 'Make Poverty History' campaign in the UK, which has developed stronger coordination mechanisms among development and non-development NGOs, and other organizations in UK civil society, and the development of much more sophisticated advocacy campaigns on aid, debt and trade. 44 CAN NGOS MAKE A DIFFERENCE? If one believes that there is a credible chain of logic linking these three conferences, their outputs, and those of other similar efforts that were ongoing during the same period, with the emergence of a more thoughtful and professional development NGO sector, and (going one stage further) linking the emergence of that sector with at least the possibility of a greater aggregate impact on development, then one can begin to answer the question posed by this volume in the affirmative, breaking down those answers by country context, type of organization, type of impact, longevity, sector, issue and so on in the ways that other chapters try to do. I think one would have to argue an extreme version of the counter (act ual to say otherwise - in other words, to claim that the world would be a better place without the rise of development NGOs, however patchy their impact may have been, especially given the huge and complex challenges that face all NGOs in their work today. Perhaps I am not setting the bar very high in making this point, but in critiques of NGOs it is often forgotten. There has been a positive change in the distribution of opportunities to participate in development debates and in democracy more broadly, and in the capacities and connections required by NGOs to play their roles effectively, even if global trends in poverty and power relations, inequality, environmental degradation and violence are not all heading in a positive direction. In other words, some of the preconditions, or foundations, for progress are being laid, brick by brick, organization by organization, community by community, vote by vote. If one believes that democratic theory works, then, over time, mote transparency, greater accountability and stronger capacities for monitoring will feed through into deeper changes in systems and structures. Civil society may yet fulfil Kofi Annan's prediction as the 'new superpower* — a statement that was largely rhetorical but contained at least a grain of truth. And as context for that conclusion, think back thirteen years to the first Manchester Conference when NGOs were still something of a backwater in international affairs. No one could say the same thing today. Where We Were Wrong, and Why It Is Important So, so far, so good. There was one major area, however, in which the analysis of previous conferences was seriously awry, and it has some significant consequences for the NGO world going forward. This was the j prediction that foreign aid would be replaced by a different, healthier and / more effective system of international cooperation in which the drivers of ' development and change would no longer be based around North-South transfers and foreign intervention. In fact, the clear decline in real aid flows that was observed between 1992 r _ t tcsc for Comfort? NGOs, States and Donors, St. Martins Press, London, pp. 257-74. Valdeŕrama, M. (2004) 'Empoderamiento y participáciou de la sociedad civil en la cooperación internacionál: el caso peruano', in F. Negrón et al. (eds), Mito y realidad de la ayuda externa: America Latina al 2004, ALOP, Lima. References Ballon, E., and M. Valdeŕrama (2004) 'Las relaciones de las ONGD de America Latina y las agencies privadas de cooperación internacionál europeas en el contexto de la globalizacion', in F. Negrón et al, (eds), Mito y realidad de la ayuda externa: America Latina al 2004, ALOP, Lima. ALAN THOMAS 91 Whatever Happened to Reciprocity? Implications of Donor Emphasis on 'Voice* and 'Impact' as Rationales for Working with NGOs in Development Alan Thomas Eliminating world poverty is a job for everyone, not just governments. In 2005, people around the world raised their voices to demand change.... NGOs will help deliver services, especially in fragile states. ... civil society groups will hold the Government to account in the UK, and encourage their counterparts in developing countries to do the same. (UK White Paper on Eliminating World Poverty, DFID, 2006: 81). This chapter concerns non-goveramental organizations and the rationale for their involvement in development. It analyses how donors view NGOs, looking particularly at the example of the UK Department for International Development (DFID), arguing that NGOs are expected to conform to one of two prescribed models of what they do, which tends to ignore or downplay the value basis of what NGOs are and the variety of ways they relate to development.1 The chapter suggests reciprocity (Polanyi, 1957) as an organizing principle that incorporates the variety of values underlying NGOs and differentiates them from both private firms, based on a rationale of self-interest and exchange through the market, and government agencies, based on a rationale of legitimate authority and coercive redistribution. At the same time, it seeks to place NGOs within 'civil society', which in political rather than economic discourse has also been used to describe the space between the state and the market. However, usage differs as to whether 'NGO' is a synonym for 'civil society organization' (CSO) or refers to one particular type of CSO — for example, one that delivers humanitarian relief or promotes 'development' for others. Both the private and state sectors are modern sectors contrasting with a 'traditional', 'community' sector, based on a rationale of mutuality, recipro-■i- j —ia„ ktoOs ran be regarded as belonging to a third modern sector, based on some of the positive values of community but with more openness and universality. Arguably this third sector also corresponds to the organizational dimension of civil society. Invoking the idea of 'civil society' is one way of investing the third (modern) sector with some positive attributes. Many authors agree that it should not be defined as just a residual category (non-profit and nongovernmental) but consists of 'value-based' or 'value-led' organizations (Paton, 1991; Hudson, 1995), though which values are to the fore is subject to much debate. Suggestions include voluntary association (Streeck and Schmitter, 1985), charity (Butler and Wilson, 1990), membership (Stryjan, 1989) , trust and solidarity (Gherardi and Masiero, 1990), enthusiasm (Bishop and Hoggett, 1986), among others. The values underlying development NGOs in particular are if anything even more varied, although many relate to participation or empowerment. Some derive specifically from movements based in developing countries, for example Freire's (1972) conscientization, or Gandhian concepts such as gram swaraj (village self-rule) or sarvodaya (the welfare of all). Other value-based ideas taken up by many NGOs, while of Northern derivation, are specific to attempts to deal with problems of development, such as Schumacher's (1973) 'small is beautiful', Korten's (e.g. 1990) 'people-centred development' and Chambers's (e.g. 1997) ideas of participative rural appraisal and power reversals. It might appear that the values involved are too diverse to generalize about the underlying principles. Some are the values of groups set up for the mutual benefit of their members while others relate to organizations set up for the benefit of others or for general public benefit. However, over time successful voluntary organizations tend to combine elements of all three categories of benefit (Handy, 1988). Indeed, all organized voluntary action can be seen as combining the human impulse to act directly in response to a perceived need with the need to pool resources by acting in groups. I suggest that the best attempt at defining this impulse in terms of a single principle is Polanyi's (i957) idea of reciprocity, where goods, services or effort are given freely not for immediate exchange but in the expectation of reciprocal assistance being available when required (a similar notion underlies Titmuss's (1970) 'gift relationship'). However, a general understanding of voluntary, non-profit or 'civil society' organizations must also recognize that they are often small and specific in their area of operation. Thus the third sector - or 'civil society organizations', including NGOs — comprises organizations which may all be value-based and rely on reciprocity but are based on a variety of specific values and focus on the needs and interests of particular groups. NGOs have become increasingly important in development since the 1980s, as the neoliberal combination of market economics and liberal democratic politics became dominant. As Edwards and Hulme explain, 92 CAN NGOS MAKE A DIFFERENCE? NGOs fitted into the 'New Policy Agenda' promoted by donors, appearing simultaneously 'as market-based actors' and 'as components of "civil society" ' (1995Y 849). Thus, on the one hand, the increase in provision of services or 'gap-filling' (Vivian, 1094) by NGOs was seen as part and parcel of the privatization of state services, despite NGOs' non-profit basis. On the other hand, NGOs were seen as prime agents of democratization (Clark, 2991), or even as intrinsically democratic simply by virtue of being part of civil society (ROAPE, 1993). In practice the contribution of NGOs to development is enormously varied and multidimensional, reflecting their sheer numbers and diversity. There is a huge difference between international NGOs, mostly based in the developed world, and indigenous local or national NGOs in the developing world. Often started as charitable relief or missionary welfare organizations, the former generally work in developing countries through their own branches or with local partner organizations, often NGOs themselves. The majority of the latter are small, but they include organizations such as the Bangladesh Rural Advancement Committee (BRAC), the largest national NGO in the developing world, with over 97,000 employees in 2005.2 BRAC and other large NGOs (especially in South Asia) often function as para-governmental or quasi-governmental organizations, operating in parallel with the state and complementing it in the provision of social services. However, for some time, many working in NGOs have wished to go beyond simply providing relief or other services within the neoliberal model of market-led development. A symposium on 'Development Alternatives: The Challenge for NGOs' held in London in March 1987 explored the suggestion of a distinctive 'NGO approach' to development based on empowerment and the idea that poor people could be supported to become the agents of their own development (World Development, 1987; see also Poulton and Harris, 1988; Thomas, 1992). However, despite a number of well-reported success stories at the local level, it was unclear whether this 'NGO approach' could have a broader impact. In one of the papers from that London conference, Sheldon Annis (1987) asked, 'Can Small-scale Development be a Large-scale Policy?', and this question of how to 'scale up' from local experience became perhaps the most important of a number of distinct challenges to development NGOs which remain relevant today. A number of writers have seen these challenges in terms of a sequence of strategies. At the same conference, David Korten distinguished between three 'generations' of NGO strategies: the first committed to relief and welfare activities, the second promoting small-scale local development that empowered local communities and broke their dependency on humanitarian > a-;-a ;„„„i,7M in a range of activities designed to achieve ALAN THOMAS 93 institutional and policy change. Later, he suggested the need for a 'fourth generation' strategy, committed to increasingly complex networks and to advocacy at international as welt as national level (Korten, 1990: 123-4). Individual NGOs could be involved in various mixes of the strategies. In a similar vein, Alan Fowler (1997: 220-21) characterized NGO activities as a mixture of three types of effort: 'welfare and delivery (the global soup kitchen)', 'strengthening people's organizations and movements', and 'learning for leverage'. He suggested NGOs should shift away from the first by either 'concentrating on building people's capacities to look after and demand tor themselves' or 'gaining leverage on structural changes to governments and markets which benefit the poor' (Fowler, 1997; 220-21). The rest of this chapter concentrates not on the NGO perspective but on how donors justify working with NGOs. The next section charts the changes in donor funding and expectations of NGOs from the 1970s to date. The following two sections analyse more closely how 'voice' and 'impact' are currently the dominant rationales put forward by donors for working with NGOs, looking in particular at policy and other statements by DFID. The final section considers how these two rationales may 'squeeze out' fundamental aspects of NGO work in development, many of which can be summed up in terms of the concept of 'reciprocity', and concludes with some implications. Changes in Donor Funding of NGOs and Its Rationale Throughout the period of the above-mentioned discussions on how to move from small-scale successes to making a bigger difference, resources for development through NGOs have increased consistently. From 1970 to 1999, NGO aid went up from US$3.6 billion to US$12.4 billion annually, equivalent to 21.6 per cent of total development assistance from members i of the OECD (see Table 5.1). For most of that time official donor grants to NGOs also increased. Although the proportion of official aid going through NGOs has reduced since the mid-1990s, private funding of NGOs continues to increase and more than offsets this decline. In fact, the proportion of NGOs' resources coming from private sources has never fallen below 65 per cent and by 1999 it was above 85 per cent and rising. Nevertheless, access to official aid funds has become extremely important to NGOs generally, and particularly for some NGOs. Thus, although NGOs have their own agendas and cannot be regarded simply as vehicles for implementing official aid policies and programmes, donors' expectations of what NGOs should do has a considerable influence on them. 94 CAN NGOS MAKE A DIFFERENCE? ALAN THOMAS 95 NGO and official aid to developing countries (constant 1990 $bn) Total NGO aid to developing countries private donations official grants OECD official aid NGO aid as % of OECD aid 1970 1980 1988 1999 3.6 5-2 6.9 12.4 3-5 3-6 4-5 10.7 0.1 1.6 2.4 1.7 29.5 42.1 51-4 46.6 ii.o 11.4 12.3 21.6 Source: Clark, 2003: 130. Within the general upward trend, official funding has been affected by contradictory factors at different times, stemming from changing donor views on how to achieve aid effectiveness and the best role for NGOs. Thus there was a dramatic increase in official aid channelled through NGOs from the mid-1970s, consolidated through the 1980s and early 1990s, influenced by the rise of governments in the West committed to neoliberal economics ! and the disenchantment of many Western donors with the performance of f government in the developing world. However, from the early 1990s most i leading donors reduced aid relative to their GNP until, by 1997, OECD donors gave the smallest share of their GNPs in aid since comparable statistics began in the 1950s - less than 0.25 per cent (World Bank, 1098: 2). Aid channelled through NGOs also fell dramatically. Since 1997, with increasing commitment to the International Development Targets (IDTs) and now the Millennium Development Goals, aid/GNP ratios are increasing again. However, donors seem to be continuing to reduce aid flows through Northern NGOs, in relative if not in absolute terms, perhaps because they are revising their view of the state as an obstacle to pro-poor change and are now working to strengthen state capacity. Nevertheless, as already noted, the resources of Northern NGOs have remained buoyant due to growth in private donations (Table 5.1). The figures in Table 5.1 are heavily influenced by trends in the United States, which by virtue of its size accounts for almost half of official and private funding channelled by Northern NGOs to developing countries. However, the various members of the 0£CD vary considerably in how much official aid goes through NGOs and how much the latter depend on these funds compared to private donations. In Britain, for example, ■„ ^^five)v jOWj although it has fluctuated over the years, increasing from 1.3 per cent between 1983 and 1986 to 7.6 per cent of DFID expenditure in 1999, and falling back to 5.5 per cent of an increased DFID budget by 2003. By comparison, the US percentage has been close to 10 per cent throughout. It is US development NGOs that are largely responsible for the statistic that Northern NGOs derive a large and increasing proportion of their funding from non-government sources, since they are even more heavily privately funded than those based in other Northern countries. Nevertheless, official aid accounts for much less than half the funds of British NGOs. According to one estimate, NGOs with an international development remit receive 20 per cent of all donations to UK charities (Randell and German, 1999a: 236), equivalent to ^1.5 billion in 2001/02. These changes in aid funding have each been accompanied by changes in donor expectations of NGOs. Thus in the 1980s, with donors favouring structural adjustment lending including deregulation, liberalization and privatization, the increase in official aid funds going through NGOs corresponded with the view that they could deliver humanitarian relief and local development effectively, reaching the poorest communities at relatively low cost. However, many Northern NGOs continued with building long-term relations of trust with Southern partner organizations and working politically towards social transformation and alternative models of development based on empowerment and reciprocity. The tension between these approaches was exacerbated with the adoption of the IDTs in the 1990s and then the MDGs. Bebbington (2005) examines the case of Dutch aid and changes in the 'co-financing programme* with Dutch NGOs up to early 2002. He shows how the need to demonstrate impact in terms of poverty reduction and other specific targets has undermined trust and partnership relationships. Dutch NGOs have shifted to working with different types of local CSO and reduced those programmes which had less immediately measurable impact, such as research or broader political empowerment through social movement organizations. More recently, along with rediscovering the importance of the state (World Bank, 1997), donors have discovered 'civil society'. There is a new rationale for working with NGOs which is applied to working with civil society organizations (CSOs) more broadly. It is argued that they can facilitate a certain type of empowerment process involving making the voice of the poor heard, thus helping to hold government agencies to account, and these 'voice and accountability' roles can help ensure that pro-poor policies are designed and implemented. Hence a partnership with CSOs that play these roles can complement a shift to the promotion of poverty reduction strategy programmes (PRSPs) together with direct budget support or sector-wide approaches (SWAps) on the part of donors.